Calls for an extended evolutionary synthesis are flourishing in the scientific literature (as highlighted by a recent Review article in this journal (Beyond DNA: integrating exclusive inheritance into an extended theory of evolution. Nature Reviews Genetics 12, 475–486 (2011)1) and Refs 2, 3), so the identification of the necessary 'extensions' has become a crucial goal to many evolutionary biologists. Here we wish to emphasize two elements: the participation of symbionts in their hosts' adaptation and the potential for evolutionary conflict between the different agents of biological information that affect the phenotype of an individual, including symbionts.

In insects, bacterial symbionts can improve host fitness, and recent studies have shown that these symbionts respond to the selection on their host's phenotype by increasing in frequency, hence permitting the adaptation of the host4,5,6,7,8. For example, within a few years, a strain of a maternally transmitted bacterium belonging to the Spiroplasma genus has invaded many North American populations of its host, Drosophila neotestacea, because of the protection it provides against a parasitic nematode6. The host has thus become resistant to the parasite owing to the presence of the symbiont, which is now part of the host's extended genome.

Beneficial, vertically transmitted symbionts can be found in numerous taxa. In addition to the well-described interactions between insects and their bacterial symbionts (see above), the adaptation of host populations based on the spread of beneficial symbionts has also been shown in plants that are defended by symbiotic fungi9. Similar processes probably also involve viruses — which can be beneficial to plant, bacterial, vertebrate and invertebrate hosts10 — and the bacterial communities living on and in vertebrates such as humans (for example, Ref. 11). The strongest evidence for the importance of symbionts as sources of evolutionary innovation probably remains the great prevalence of mitochondria and chloroplasts in eukaryotes and the symbiotic origin of these organelles12.

The presence within the same organism of different sources of biological information (for example, nuclear genes and symbionts) that have different effects on the phenotype and different modes of transmission may also lead to evolutionary conflicts rather than favouring adaptation. Again, such conflicts are well-described in insect–bacterium associations. Indeed, several bacteria that have recently been shown to be involved in host adaptation were previously known as parasitic manipulators of sex determinism and reproduction (for example, Refs 5, 6). Conflicts may even occur when components of inclusive inheritance are transmitted in the same manner. For example, Danchin et al. discuss the differential imprinting of maternally and paternally derived genes1 as a component of inclusive inheritance. Actually, conflicts between nuclear genes controlling imprinting in both parents and the offspring are the leading hypothesis for the evolutionary origin of such imprinting13,14. An inclusive understanding of the different types of biological information and their influence on phenotypic evolution thus necessitates the simultaneous consideration of these different elements and their potential conflicts.